Wide-angle scattering studies of enamel

It has been known for many years (see Miles 1967) that the lattice parameters of dental enamel, particularly a, differ significantly from those of stoichiometric HAP. For example, for human dental enamel, a = 9.441(2) A and c = 6.878(1) A for a fraction with density 2.95 g cm (Young and Mackie 1980), and a = 9.4555(76) A and c = 6.8809(47) A (Wilson et al. 1999) for pooled enamel from permanent teeth have been reported. Both sets of measurements used Rietveld methods applied to XRD data. These values compare with a = 9.4243(55) A and c = 6.8856(35) A (Morgan et al. 2000) for stoichiometric HAP, also determined from Rietveld analysis. Thus the a-axis parameter for dental enamel is about 0.3% larger than for HAP. 

X-ray diffraction patterns from enamel sections cut in the longitudinal direction of a tooth show that the c-axes (the hexagonal axes) of the apatite crystals (also their long direction) are highly oriented in the direction of the enamel rods (sometimes called prisms) (Miles 1967). The rods run from the enamel surface (to which they are nearly normal) to the enamel-dentin junction and are about 4-7 pm in diameter. 

Glas and Omnell (1960) made use of the high degree of preferred orientation of the apatite crystals in hippopotamus enamel to measure crystal dimensions over several orders of diffraction. They found that the crystal dimension in the a-axis direction (the width of the crystals) from the Scherrer formula (410 A, approximate precision limits 370 and 450 A) was independent of the order of diffraction. However, dimensions in the c-axis direction (the length of the crystals) calculated from the 002, 004, 006, and 008 peaks were 741 100, 466 40, 361 35 and 295 25 A respectively. After correction for 

Lattice parameters calculated from wide-angle XRD patterns of human dental enamel showed no measurable shift when the F content increased from 70 to 670 parts per million (Frazier 1967). However, the 002, 211, 300 and 202 peaks became slightly sharper. The Scherrer formula showed increases in size from 1420 100 to 2740 240, 780 25 to 1030 25 and from 780 20 to 1000 20 A from the 002, 300 and 211 peaks respectively. In the absence of further study, it is not possible to say whether these changes were due to a real increase in size or to a reduction in microstrain. 

The broad diffraction lines of the mineral in bone make accurate measurement of lattice parameters problematical, hence such measurements are rarely attempted. Nevertheless, Handschin and Stern (1992) measured the lattice parameters of human iliac crest samples of 87 individuals aged 0-90 years. Determinations were based on a weighted least-squares analysis of the positions of 6 lines (002, 102, 310, 222, 213, 004) and gave average statistical errors of 0.002 A in a and 0.003 A in c. The parameters showed a slight reduction with age (a by 0.00015 A and c by 0.00005 A, both per year). Chemical analyses of such samples have been reported (Handschin and Stern 1994 and 1995). It may be that the reduction in a-axis is in part caused by the increase in F content with age noted earlier (Wix and Mohamedally 1980). 

---